Constant Amplitude Zero Auto-Correlation

Description

CAZAC sequences are a class of complex-valued sequences characterized by two key mathematical properties: constant envelope (amplitude) in the time domain and perfect periodic auto-correlation, meaning the correlation of the sequence with a time-shifted version of itself is zero for all non-zero lags. The most prominent CAZAC sequences used in 3GPP standards are Zadoff-Chu (ZC) sequences, named after their inventors. A Zadoff-Chu sequence is defined by a root index 'u' and a length 'N', where N is a prime number. The sequence exhibits constant amplitude, ensuring a low Peak-to-Average Power Ratio (PAPR), which is beneficial for power amplifier efficiency. Its zero auto-correlation property means that when the sequence is correlated with a cyclically shifted version of itself, the result is zero unless the shift is zero, providing ideal orthogonality between different cyclic shifts of the same root sequence.

In the 3GPP LTE and 5G NR physical layer architecture, CAZAC sequences are utilized to generate several critical signals. Primary Synchronization Signals (PSS) and Secondary Synchronization Signals (SSS) for cell search and timing acquisition are derived from Zadoff-Chu sequences. Uplink Demodulation Reference Signals (DM-RS) and Sounding Reference Signals (SRS) are also constructed using these sequences to facilitate accurate channel estimation at the base station. Furthermore, the Physical Random Access Channel (PRACH) preamble is generated from a Zadoff-Chu sequence, where different root indices and cyclic shifts define orthogonal preambles for random access attempts, minimizing collision probability.

The generation and processing of CAZAC sequences involve specific mathematical operations defined in the 3GPP specifications. The base station and user equipment (UE) generate these sequences locally based on known formulas and parameters (root index, sequence length, cyclic shift). For channel estimation, the receiver correlates the received reference signal (which is a known CAZAC sequence distorted by the channel) with the locally generated clean sequence. Due to the zero auto-correlation property, this correlation effectively isolates the channel's impulse response, allowing for precise estimation of channel delay spread and frequency response. This is crucial for coherent demodulation of data symbols in OFDM/OFDMA systems.

The role of CAZAC sequences extends to providing orthogonality in both time and frequency domains. Different UEs can be assigned different cyclic shifts of the same Zadoff-Chu root sequence for their reference signals, creating orthogonal resources that do not interfere with each other even when transmitted simultaneously. This orthogonality is vital for uplink multi-user MIMO and for accurate channel estimation in high-mobility scenarios. The constant amplitude property also simplifies transmitter design and improves power amplifier efficiency, as it reduces signal distortion and out-of-band emissions.

In summary, CAZAC sequences are a foundational mathematical tool in modern cellular systems. Their ideal correlation properties underpin critical physical layer procedures including synchronization, channel estimation, and random access. The specifications detailing their application, such as TS 36.211 for LTE and TS 38.211 for NR, define the exact mapping of these sequences to time-frequency resources, ensuring interoperability and robust performance across different network implementations and device vendors.

Purpose & Motivation

CAZAC sequences were introduced to solve fundamental challenges in Orthogonal Frequency Division Multiplexing (OFDM)-based wireless systems like LTE and NR. Prior approaches for reference signals and synchronization used sequences with less ideal correlation properties, which could lead to higher channel estimation errors, increased interference between users, and higher Peak-to-Average Power Ratio (PAPR). High PAPR is particularly problematic for OFDM as it reduces power amplifier efficiency and increases cost and power consumption in user devices. The need for a sequence with perfect periodic auto-correlation and constant amplitude was driven by the requirement for highly accurate, low-complexity signal processing in the physical layer.

The primary problem CAZAC sequences address is the need for a 'perfect' reference signal for channel estimation. In a multipath fading channel, the receiver must estimate the channel's frequency response to correctly demodulate data. A sequence with perfect auto-correlation ensures that when the received reference signal is correlated with the known transmitted sequence, the result is the channel's impulse response without any interference from the sequence itself. This leads to minimal estimation error. Furthermore, the constant amplitude property directly combats the high PAPR inherent to multi-carrier OFDM signals, allowing for more efficient and linear power amplification.

Historically, before the adoption of CAZAC sequences in 3GPP LTE (Rel-8), other sequence types like pseudo-random noise (PN) sequences were used in systems like WCDMA. While PN sequences have good cross-correlation properties, they do not possess the perfect periodic auto-correlation or constant amplitude of CAZAC sequences. The shift to OFDMA for the LTE uplink (SC-FDMA) and downlink necessitated reference signals that were orthogonal in both time and frequency domains and friendly to power amplifiers. The mathematical properties of Zadoff-Chu sequences made them the ideal choice, enabling the high spectral efficiency, robust mobility performance, and device power efficiency that are hallmarks of 4G and 5G systems. Their adoption was a key enabler for the performance leap from 3G to 4G.

Key Features

• Constant amplitude in the time domain, leading to low Peak-to-Average Power Ratio (PAPR)
• Perfect periodic auto-correlation (zero for non-zero lags) for precise channel estimation
• Generation of orthogonal signals through different root indices and cyclic shifts
• Foundation for critical physical layer signals (PSS, SSS, DM-RS, SRS, PRACH)
• Robust performance in high-mobility and multipath fading environments
• Enables efficient implementation of uplink multi-user MIMO through orthogonal reference signals

Evolution Across Releases

Defining Specifications